Qualitative Cognition for Uncertainty Knowledge Using Cloud Model

Concepts are basic elements of natural language processing, studying on concept representation and transformation between connotation and extension become more and more important. Multi-granularity concept extraction is still a difficult problem in uncertainty knowledge representation. Cloud model is an uncertainty cognition model, which realizes the bidirectional transformation between a qualitative concept and quantitative data by Gaussian cloud algorithm. Gaussian cloud transformation provides a method to transform a group of data in problem domain to multiple concepts in different granularities in cognition domain. This paper introduces cloud model and Gaussian cloud transformation algorithm to describe the multi-granularity concepts. A case study is also given to prove the effectiveness of the proposed method.